MM5823 Frequency Dividers

[germaniac]

Hi Friends,

I just got an MG-1 and so far it's a mixed blessing. My nerves were steeled to get inside and clean out the notorious Black Muck, and lo and behold, there wasn't any! At some point in this unit's lifetime, somebody took out the foam before it got bad. There wasn't even any post-cleanup residue. For anyone who's ever undertaken this task (I did an Opus 3), you know this is worth about two hours of your time, not to mention a lot of Q-Tips.

So much for the good luck. On the down side, it seems one of my MM5823 divider chips is going south. At first I thought it might be some of the trigger buss contacts because I could get the dead notes to sound by keeping a lower key pressed. But on closer study of the schemo, I found that all the triggers are extracted AFTER the polyphony summer. This means anything that goes wrong with those divider chips will affect the behavior of the whole keyboard.

So, here come the questions: What is the deal with these chips failing? Why do they go bad? Is it the chip itself that's unreliable or is it a design problem? By design problem, I mean I wonder if there's some sort of circuit modification that might help protect these chips from future damage. If not, it seems like a losing battle with any keybaord that uses the MM5823, because eventually they will all fail, and eventually there won't be any left to replace them with.

Anyone have thoughts/ideas/suggestions on this? Thanks!

Regards,
Joe

[electrical_engineer_gEEk]

be THANKFUL that you have Chipsforbrains on ebay selling them recently.....
it took me forever and alot of money to track down two of those chips a few years ago......

remember to give Chipsforbrains a good pat-on-the-back for providing all us restoration geeks supplies...otherwise my mg-1 & opus would be sitting in a trash-bin!



As far as having any experience with the chip? I have no clue, they could be more susceptible to voltage spikes, maybe they don't age well, maybe they weren't manufactured well...maybe all of the above...I'm not sure, sorry

best of luck with your restoration!

[MC]

They're also used in Liberations and I don't hear any hollering about 5823s failing in those. It is possible they are sensitive to power rails and that the MG-1 does not have as robust a power supply as the Liberation.

[germaniac]

Hi EEG: If I go the chipforbrains route I'll be sure to thank him as you said.

Hi MC: Thanks for your input about the Liberations. This tilts the scale a little towards evidence of a design flaw in the MG-1. If the chip is sensitive, maybe some current protection resistors on the power rails and some extra decoupling/filtering might help. Any other ideas from those of you with better electronic chops than mine welcome!

Joe

[Kevin Lightner]

I just sold a bunch of these chips on Ebay. Where were you?

The 5823 is one of those rare logic ICs that operates on 30 volts like many opamps.
Unlike opamps though, they have no short circuit protection at all.
If something happens to an output pin, they can internally short and die.
Possibly aging filter caps play a part too.

I don't think the Liberations are any more reliable than the MG-1.
The fact is, the Liberation didn't sell well and the MG-1 sold like hotcakes.
So logically there would be more number of failures for MG-1s.
Also MG-1's saw more abuse overall. They were cheap, easy to move synths.
They didn't cost over a thousand and come with their own hardshell case.
The Polymoog and Opus have seen about the same percentage of failures in my opinion.

[germaniac]

Hi Kevin,

Okay, well, what you say about the reliability of the Liberations and others makes sense, and tilts the scale of blame back again to the chip. I wonder what short-circuit protection on one of those chips would entail--diodes on every divider output?

As far as your Ebay offer, I only got the MG-1 last Saturday and didn't know what the diagnosis was at that point. Right now, I'm working with a friend (Scott Stites) on cooking up a retro-fit with CMOS counters, but regardless if that pans out, is there any chance you still have a chip or two to spare?

Regards,
Joe

[MC]

No, diodes will protect against static discharge and overvoltage but not short circuits. A series resistor on an output will protect against short circuits but it has to be a value large enough to limit current to a safe value and to avoid impacting the output signal integrity at the same time.

[Kevin Lightner]

Not sure on the resistors.
Too high a value and it possibly won't work.
Too low and it won't offer much protection.
They don't source much current, so resistors on their power pins might be an easy cheat.
Diodes on the outputs might be ok, but would you really want to add one for every note?
Also, the odds of causing an unnoticed short or static zap might actually destroy chips in the name of saving them.

For all we know, these chips might fail because another chip failed at one time and took down a power rail. I have no idea what these dividers experience when powered up lopsided.
One bad connector or loose fuse (in a power supply that has separate fuses) can cause this sort of condition.
Making sure the grounds and power rails are good is important.

[germaniac]

They don't source much current, so resistors on their power pins might be an easy cheat.

That sounds do-able. Something in the 100-ohm range maybe?

For all we know, these chips might fail because another chip failed at one time and took down a power rail. I have no idea what these dividers experience when powered up lopsided. One bad connector or loose fuse (in a power supply that has separate fuses) can cause this sort of condition.
Making sure the grounds and power rails are good is important.

Well, if I can get lucky with the divider fix, I'm definitely going to replace the PS filter caps, possibly the regulators, and add some overall PS decoupling to the sensitive chips before I fire it up again.

Thanks Kevin.

Regards,
Joe

[electrical_engineer_gEEk]

yeah, i think we really luck out because most of the power supply chips and caps are still available these days so they are easy to obtain and most likely will last alot longer then the originals. I'm so glad they've kept packages like TO220s and DIPs (not to mention through hole electrolytic caps) around because it at least gives us a fighting chance for drop in replacements!

but alas, the whole world may go surface mount soon ( i know my company has)....what will happen to hobbyists and proto boards, will they be around forever?

[MC]

MOTM is going SMT and is no longer selling kits.

More and more DIP parts are disappearing from the catalogs I'm afraid.

Even the Buchla 200e modular is SMT (I got a peek at the filter module).

While the Voyager is largely DIP (digital board being the exception), the LP is SMT.

[Kevin Lightner]

Considering that it's still possible to find fairly rare and specialized chips made twenty five years ago, standard parts like 78XX type regulators will probably still be available for awhile.
There's many devices made with through-hole components and the fact that you expresss concern for their future here shows that others are thinking the same way too.
The demand will just be too great to ignore and smaller manufacturers and surplus places will be selling them if they can.
I also feel there will be greater interest in SMD to DIP carriers and adaptors and their prices will drop. This will allow some SMD packages to be used in earlier products for repairs.
Some things may not change much at all. It's difficult to get the capacity and current rating of larger electrolytic cap down to SMD size. I'm sure they're working on it though.